The night sky images from Mauna Kea are being taken continuously using
the NSF-supported Concam. These images are at http://www.concam.net.

In preparation for running in subsequent years, we want to start
archiving these images (in FITS format) so we can compare our
cloudiness diagnostics with other sources (weather maps, water vapor
meaurements from the sub-mm telescopes on Mauna Kea, etc.). The archive
is at http://concam.net/mk/archive.html

Furthermore, we want to develop an algorithm to determine the sky
transparency. Since the moon causes ghost images, and the images are in
the optical, detecting dark regions due to clouds will be difficult.
Rather, since known stars can be located, they can be photometered and
the values compared within and between nights. Note that the image
quality deteriorates at the edges of the field, so the photometry will
need to allow for this.

To locate the stars requires knowledge of the time at which the image
was taken.  Formally one also needs to know the latitude and longitude
of the observatory, the altitude and azimuth of the "telescope" and the
astrometric transformation. In reality, an hour angle, declination map
onto detector pixels should be sufficient.

Ultimately the photometry information needs to be fed to observing
scheduler. It should also be archived so that the reduction pipeline
can use this information to confirm extinction trends from the SNIFS
imager.